SDSU Financial Markets Lab · 2026-27

AI is real.
The price is a bet.
We measure the gap.

A year-long undergraduate research project. We take today's AI stock prices, work backward to what they quietly assume about 2030, and test whether all those assumptions can be true at the same time. Starts September 17. No experience needed.

The test at the center of the project
What every AI bet implies for 2030 WE COMPUTE THIS
The biggest pie it could come from WE MEASURE THIS

If the first bar ends up bigger than the second, they cannot all be right. That is a fact about today, not a prediction about tomorrow.

$0B
What OpenAI plans to spend on compute through 2030
Reported plan
$0B
What OpenAI was valued at in its last private round
March 2026
0%
Share of the S&P 500 now tied to the AI value chain
JPMorgan Asset Management, Sept 2026

The Question

How much of AI's price
is already real?

Nobody in this lab thinks AI is fake. It is real and it is making people more productive. But the numbers being thrown around right now are unlike anything we have seen, and somebody should actually measure them.

So we are not asking whether AI is good or bad, and we are not calling a bubble. We are asking something much narrower: how much of today's prices is backed by cash and output that already exist, and how much is a bet on things that have not happened yet?

Doctors do not predict the day you will have a heart attack. They measure your blood pressure and tell you what it means.

That is the whole design. Every claim we make is a measurement about right now, true or false the day we compute it, no matter what the market does next. Ask "is this a bubble" and you cannot know for years. Ask "what has to be true for this price to make sense" and you can answer today.

Last season · completed

Forecasting the FOMC

A multinomial logistic regression trained on 163 Fed meetings (1998 to 2025), forecasting each decision as a hike, hold or cut from the macroeconomic data available beforehand. Built end to end by students.

RBloombergClassification28 yrs

Why Join

What you leave with.

A paper with your name on it

Real published research for your resume or grad school application.

→

Valuation you can actually do

Build a DCF, then run it backward. This is the skill the interview asks about.

→

Bloomberg Terminal time

Plus R for the modeling. Real market data, not textbook examples.

→

Your own company to cover

One name, all year, yours. You become the person in the room who knows it best.

→

One hour a week

Built to fit around a full course load.

→

Join the Lab Now recruiting

Take a seat.

Open to any SDSU student, any major, any year. You do not need to have taken econometrics, and you do not need to know R. Plenty of people arrive knowing neither.

First session is Thursday, September 17, and the project runs through April 22. Fill out the form and we will send you the room details and what to read before it.

Takes 30 seconds · no commitment yet

Sign upSDSU FML · 2026-27

The form asks four quick things:

  • Your name and year
  • Your SDSU email
  • Your major
  • Anything you want us to know
Open the sign up form →

Opens in Google Forms · takes about 30 seconds

Discord Get in the room before day one. Meeting reminders, the reading, and everyone already in the lab. Join the server →

The Fine Print

Here's the info.

When
Thursdays, 6:30 to 7:30pm
Sept 17 to April 22
Where
Love Library, room 261
Who
Any SDSU student, any major
Stack
R, Python and the Bloomberg Terminal

Want us to visit your class? mkocherga7657@sdsu.edu

Got a question? Ask it here and we will get back to you.
Got it. We will reply to that address shortly.